Detection Of Lunar Impact Flashes Using Maximum-Based Image Stacking

Abstract

The detection of meteoroid impacts on the lunar surface provides critical constraints on the meteoroid flux in the Earth–Moon system and informs risk assessments for future lunar missions. Currently used detection software packages in ground-based lunar impact research primarily rely on analysis of individual image pixels, searching for statistically significant deviations in space or time domain that can be associated with Lunar Impact Flashes (LIFs). In this work, we present an alternative detection method based on maximum-value image stacking, very well suited to highlight impulsive, localized brightness increases associated with lunar impact events. Unlike traditional averaging techniques, the maximum stacking algorithm retains peak intensity values across image sequences, enhancing the visibility of transient phenomena. We apply the method to real observational datasets recorded during a period of particularly high lunar impact rate (the Geminids meteoric shower peak on December 13–14) and demonstrate its ability to identify impact candidates. Comparative analysis shows that maximum-based stacking improves detection probability for brief and weak events while maintaining computational simplicity. This technique represents a complementary tool for lunar impact monitoring and can be readily integrated into existing detection pipelines.